Intelligent scheduling and resource prediction framework for multidimensional data streams in cloud computing environments

Cloud computing has become a foundational technology for handling the growing demands of data-intensive applications. Efficient resource allocation and scheduling are critical challenges, particularly when managing multidimensional data streams in dynamic environments. This paper introduces an intelligent framework for scheduling and resource prediction tailored to cloud computing systems. The proposed method integrates advanced machine learning techniques to predict resource demands accurately, enabling dynamic scheduling that optimizes resource utilization while reducing delays. The framework also includes an energy-aware optimization module that minimizes power consumption without compromising performance. The methodology leverages historical and real-time data to ensure adaptability and scalability across diverse workloads. Experimental results demonstrate the framework’s high prediction accuracy, with minimal mean absolute error and root mean square error, its ability to maintain over 91% resource utilization under varying workload intensities. Moreover, the energy efficiency improvements of up to 15.4% highlight the system’s sustainability. This study contributes a robust and scalable solution to modern cloud resource management, addressing both operational and sustainability goals.